DeepChem — molecular machine learning
Run end-to-end molecular ML with DeepChem: featurizers, MoleculeNet benchmark datasets, scaffold splitting, GNN and pretrained models (ChemBERTa, GROVER) for property prediction (ADMET, toxicity, solubility)
- What
- Run end-to-end molecular ML with DeepChem: featurizers, MoleculeNet benchmark datasets, scaffold splitting, GNN and pretrained models (ChemBERTa, GROVER) for property prediction (ADMET, toxicity, solubility)
- Cost
- Free
- Needs
- a dedicated Python environment (Python <= 3.11 for DeepChem 2.8.0 stable; uv recommended) — DeepChem's dependency pins conflict with current scientific stacks, so don't install it next to pandas 3 / current PyTorch; optional GPU for GNNs; your own molecular data (SMILES, SDF)
- Install
- Copy the installer prompt below into your Muse — your agent does the rest.
Curated by Skill Harbor — @alterlab-ieu's alterlab-deepchem skill, listed here with credit to its creator (part of the AlterLab Academic Skills suite): an end-to-end reference for molecular machine learning with the DeepChem library. It covers loading chemical data (SMILES CSVs, SDF files, FASTA protein sequences), molecular featurization (circular fingerprints, RDKit/Mordred descriptors, graph featurizers) with a decision tree for picking the right featurizer, data splitting done right (scaffold splitting to prevent leakage), model selection and training (sklearn wrappers, multitask regressors, GCN/GAT/AttentiveFP graph neural networks), 30+ MoleculeNet benchmark datasets with standardized splits, and transfer learning with pretrained models (ChemBERTa, and GROVER with the honest warning that DeepChem's GroverModel is not a one-line pretrained loader). Three production-ready scripts ship with it (solubility prediction, GNN training, transfer learning), plus best-practice patterns and a pitfall-to-fix catalog (leakage, GNNs underperforming fingerprints, overfitting on small data). Honest caveats: DeepChem 2.8.0 needs Python ≤ 3.11 in a dedicated environment — its pins conflict with current scientific stacks; GPU helps for GNNs; your molecules come from you — the skill is workflows and code, not data; results are research-grade predictions, not chemistry advice for real decisions. MIT licensed. Skill Harbor never reviews the code, review it yourself before use. Discovered via skills.sh.
Version:
Install
Prerequisites: a dedicated Python environment (Python <= 3.11 for DeepChem 2.8.0 stable; uv recommended) — DeepChem's dependency pins conflict with current scientific stacks, so don't install it next to pandas 3 / current PyTorch; optional GPU for GNNs; your own molecular data (SMILES, SDF) Install "DeepChem — molecular machine learning" for me. It gives my agent @alterlab-ieu's DeepChem reference: molecular data loaders, featurization with a selection decision tree, scaffold splitting to prevent leakage, model training (sklearn wrappers, multitask models, GCN/GAT/AttentiveFP GNNs), 30+ MoleculeNet benchmark datasets, transfer learning with ChemBERTa/GROVER (with the honest caveats), three production-ready scripts (solubility prediction, GNN training, transfer learning), best-practice patterns, and a pitfall-to-fix catalog. Part of the AlterLab Academic Skills suite. MIT licensed. Repository: https://github.com/alterlab-ieu/alterlab-academic-skills/blob/main/skills/cheminformatics/alterlab-deepchem/SKILL.md 1. Fetch the SKILL.md file (and any helper files) from the repository path into a temporary folder and summarize what it does in one or two sentences. 2. Safety check: review the SKILL.md and scripts for anything suspicious (unexpected network calls, shell commands, credential harvesting). This repo should contain zero secrets in code. Verify that holds here; STOP on any red flag and tell me. 3. Install it as a skill: copy SKILL.md and its helper files into the agent's skills directory, in a folder named "alterlab-deepchem". 4. Verify with no network calls: frontmatter valid, files in place. 5. Report what was installed, where, and what I still need to do myself (e.g. create a dedicated environment: uv venv --python 3.11 && uv pip install deepchem (add [torch] for GNNs); prepare my molecular data; always scaffold-split molecular data). GitHub is optional: if I have a GitHub account or the gh CLI, you may use it; otherwise public access is fine. Never require it unless it's in the prerequisites above. Rules: don't touch anything outside the temp folder and the install target. If anything looks off, stop and ask me.
Questions
How do I install a build?
Every product page includes a copy-paste install prompt. Paste it into your Muse and it sets the build up for you — no manual configuration.
Where does my money go?
Straight to the seller. Skill Harbor never processes payments: checkout happens on the seller’s own page, usually Stripe.
What does the ✓ next to a creator’s name mean?
It means we confirmed the identity of the person behind the listing. It says nothing about the code itself — always check a build before installing it.